Probiotics: windows of opportunity

Although the majority of probiotics are used by reasonably healthy adults, it has often been suggested that probiotics may have the greatest opportunity to impact health of populations that are in a state of physiological transition or who are exposed to greater health challenges. Thus, there is great interest in how probiotics may impact infants, whose microbiome is in a state of flux for at least the first year of life, and elderly, who in addition to experiencing microbiota changes, may also have compromised immune function and increased risk from infectious diseases. Other populations that may benefit are those in developing countries, where incidence of infectious diseases is higher. Indeed, a study evaluating the impact of a probiotic combined with a prebiotic showed such success on reducing sepsis among infants born in poor regions of India that the study was prematurely stopped for ethical reasons (Panigrahi et al. 2017).

Studies evaluating the impact of probiotics in infant populations are many. Randomized controlled trials have examined probiotic effects on preventing necrotizing enterocolitis (Thomas et al. 2017), treating acute pediatric gastroenteritis (Szajewska et al. 2013), reducing incidence of common infectious diseases (Hao et al 2015), and managing symptoms of colic (Sung et al. 2018), improving feed-tolerance (Indrio et al. 2017) and preventing atopic dermatitis (Zhang et al. 2016).

Studies on the elderly are not as common. Recently a systematic review and meta-analysis was published on the impact of probiotics on incidence of infections in the elderly.  Fifteen randomized controlled trials comprising 5916 participants were included in this review, which concluded that probiotics do not reduce the infection rate or duration of infections in the elderly. No probiotic-specific safety concerns were noted. Study authors note several confounding factors that may impact susceptibility to infection in this population: seasonal variations in disease incidence, social structures, vaccination status, nutrition, frailty, immunosenescence and setting.

Many of the included studies were ranked as having a high risk of bias:

  • Some lacked clear reporting of randomization and allocation concealment
  • Some insufficiently reported blinding
  • Some were incomplete in reporting of outcome data
  • Some did not clearly eliminate the possibility of selective data reporting
  • Lack of independence from commercial funders

Taken together, this high risk of bias led to the evidence being considered low quality.

We might be tempted to conclude from this study that probiotics are not useful for the elderly. But such a conclusion is too sweeping. This review specifically excluded nosocomial infections, such as caused by C. difficile. C diff is a big problem in elderly populations and a few studies suggest that probiotics may be beneficial for reducing C diff incidence (Goldenberg et al. 2017). Further, some other outcomes for probiotic use in the elderly for which there is some evidence of benefit include prevention of AAD (Hickson et al. 2007) and improvement of bowel habits (Zaharoni et al. 2011). Further, as the review concluded, in the absence of high quality studies, it is difficult to come to a clear conclusion.

I’ll close by providing a few words about reading systematic reviews. They start by defining what the field calls “PICOS”. These factors define the scope of the review. Let’s look at the PICOS for the meta-analysis discussed herein:

  • Population: Defined for this review as age 65 or older with no signs of or diagnosis of infection. There were no restrictions on sex, comorbidities, or cognitive status. However, nosocomial infections (such as difficile) were excluded. Further, critically ill, immunosuppressed, oncology patients and post-operative patients were excluded.
  • Intervention: Any specific and identified probiotic strains (single or multiple strains) at least 107/g were included. (I find this very odd, and perhaps it is an error in the paper, but doses for probiotics are always given as per day or per dose, not per gram.) All regimes of administration timing were included, although I presume they included only oral administration. Interestingly, probiotics combined with prebiotics were not excluded.
  • Comparator: Only randomized, blinded trials of any duration using a comparator of placebo were included.
  • Outcomes: Primary outcomes were occurrence of infection, incidence of adverse events, mortality, and quality of life. Secondary outcomes were duration of infection, visits to emergency departments, occurrence of antibiotic use, duration of antibiotic treatment and hospitalization (occurrence or duration).
  • Setting: Studies of subjects living in the community, hospital or in long term care units were included. No restrictions on geographical region, and included studies were conducted in different European countries, Japan and USA.